The study proposed here addresses a major problem in stress research-the inability of stress theory, as it is currently formulated and tested, to explain social status differences in psychological distress (e.g., differences by gender, age, social class). To rectify this problem, a new theoretical category of major life stressors is suggested; these are termed "identity- relevant" events. It is proposed that the psychological impact of major life events depends on their meaning for individuals' social self-conceptions; some events threaten or disrupt highly valued role-identities (identity-threatening events), others add to or facilitate the individual's role-identity performances (identity- enhancing events), and still others are identity-irrelevant and therefore should have minor psychological impacts on individuals. It is expected that (a) the meanings of events for identity and (b) exposure to identity-relevant events will vary systematically by social status (in this study, by gender and marital status). Thus, when all else is equal, variations in identity structures and in exposure to identity-relevant events should better explain social status differences in psychological distress, compared to events as they are usually measured in the typical stressful life events approach. To test the predictive utility of this conceptualization, a two-wave prospective panel study of the stress experiences, identity structures, and psychological well-being of a stratified random sample of married (N=300) and divorced (N=300) men and women is proposed. Structured personal interviews will be conducted at both time points, three years apart. The predictive utility of an identity approach to gender and marital status differences in psychological distress will be empirically contrasted with findings using conventional measures of stressful life events, coping, and social support resources.